We propose investigation of molecular microdosimetry for biological effects due to weak and strong electric fields in multicellular structures. A new simulation method will be used to predict physical quantities: equipotentials (electric fields), transmembrane voltages, current densities and power dissipation density (SAR). The predicted physical quantities will then be used with three classes of biophysical mechanism (1) voltage-gated channels, (2) electroporation and (3) alteration of biochemical processes by local heating) to predict field- induced molecular change, molecular dose (change per cell), and exposure thresholds. As demonstrated in preliminary work, a single simulation/model can describe both weak and strong field bioelectric behavior. The sites of maximum molecular (chemical) change within multicellular structures will be estimated for weak and strong fields. Aims. We will: (1) Further develop the new bioelectric simulation method for multicellular structures, (2) Develop a molecular microdosimetry approach, (3) Estimate molecular dose for multicellular structures, and (4) Estimate exposure thresholds for various multicellular structures for weak and strong fields by quantitatively comparing molecular dose to molecular change due to other sources. Significance. Weak fields: Understanding molecular change-based thresholds for multicellular structures is a central problem in assessing possible environmental field effects at 50 - 60 Hz, RF and microwave frequencies. Strong fields: In vivo electroporation-based gene therapy, localized anticancer drug delivery, and electrical injury share a common feature with weak fields: Particular tissue regions are expected to be preferentially involved. Previous Work. We have used biophysical mechanism models, molecular change estimates and signal-to-noise ratios to estimate weak field thresholds for isolated cells and multicellular structures, and the biophysical mechanism of electroporation to estimate molecular transport and uptake for strong fields. Methods. We will use established biophysical mechanism models and molecular change signal-to-noise ratio methods. The new simulation method will be run on single CPU microprocessors and (for problems with greater computational complexity) a Beowulf computer cluster.